The equilibrium binding and covalent interactions of carcinogens and antineoplastic agents with double-stranded DNA involves complex steric and electronic factors that are not well understood. This is due in part to the ability of DNA to adopt different conformations under different conditions. An example of this is the efficient N7-G-to-N7-G interstrand crosslinking of DNA by the nitrogen mustard mechlorethamine at complementary 5'- GNC-3'/3'-CNG-5' sequences (N=any nucleotide). This crosslink is highly deformed because the connector between the two strands is greater than 1.5 Angstrom units too short to accommodate a classical B-DNA structure. We propose that there is a connection between the formation of the unanticipated 5'-GNC crosslink from mechlorethamine, which passes through a cationic monofunctional lesion, and the observed bending of DNA induced by the localization of cationic charge in the major groove by the introduction of 5-(omega-aminoalkyl)uridine residues into DNA. Specifically, it is proposed that the initial monofunctional mustard adduct induces a conformational change, i.e., kink, in DNA that allows the 5'-GNC and N7-G-N3-A crosslinks to form. The Specific Aims are to: (1) Determine the location of the omega- aminoalkyl sidechains and their affect on DNA structure using NMR, crystallography, fluorescence resonance energy transfer, and chemical probes. The impact of a charged purine, i.e., N7- methylguanine, on DNA structure will also be determined. (2) Evaluate how the phasing of cationic sidechains with an A-tract modulates DNA bending as measured by aberrant gel mobility. These phasing experiments will allow us to verify how and to what extent the cationic sidechains distort DNA. (3) Thermodynamically characterize DNA containing charged sidechains, and a N7-methylguanine lesion. The goal is to determine how the cationic charge on the sidechain and/or nucleotide induces changes in solvation that can contribute to DNA bending. (4) Explore the sequence selectivity and kinetics for DNA crosslinking by nitrogen mustards and non-charged analogues. The results from this aim will provide evidence for the role that stable and/or transient cationic charge plays in the sequence selectivity for N7-G to Ny-G interstrand crosslinking. The formation of the putative N7-G to N3-A crosslink by mechlorethamine will be also verified. Completion of these Specific Aims will provide new and fundamental information on: (a) the origin of the sequence specificity of antineoplastic DNA crosslinking agents; (b) how to design efficient DNA crosslinking agents; and (c) how the interaction of DNA with charged species, including basic amino acid sidechains, can induce DNA deformation.